Stop updating the ollama.com pages from the workflow

.github/workflows/publish-ollama.yml

@@ -1,22 +1,20 @@
 name: publish-ollama
-# Every change to main that touches a Modelfile or an Ollama page is published on ollama.com:
-# the changed models are rebuilt and pushed as Tromset/<name>, then the pages are updated from docs/ollama/.
-# Needs two repository secrets (Settings > Secrets and variables > Actions):
-#   OLLAMA_KEY      the private key ~/.ollama/id_ed25519 of a machine linked to the Tromset account (`ollama signin`)
-#   OLLAMA_API_KEY  an API key from https://ollama.com/settings/keys
+# Every change to main that touches a Modelfile is published on ollama.com:
+# the changed models are rebuilt and pushed as Tromset/<name>.
+# The model pages are pasted by hand from docs/ollama/: ollama.com refuses page edits made with an API key.
+# Needs one repository secret (Settings > Secrets and variables > Actions):
+#   OLLAMA_KEY  the private key ~/.ollama/id_ed25519 of a machine linked to the Tromset account (`ollama signin`)
 on:
   push:
     branches: [main]
     paths:
       - 'Modelfile*'
-      - 'docs/ollama/**'
       - 'scripts/publish.sh'
-      - 'scripts/update-ollama-pages.sh'
       - '.github/workflows/publish-ollama.yml'
   workflow_dispatch:
     inputs:
       models:
-        description: 'Models to publish (space-separated names, "all", or "none" for pages only)'
+        description: 'Models to publish (space-separated names, or "all")'
         default: all
 concurrency:
   group: publish-ollama
@@ -28,7 +26,6 @@ jobs:
     timeout-minutes: 120
     env:
       OLLAMA_KEY: ${{ secrets.OLLAMA_KEY }}
-      OLLAMA_API_KEY: ${{ secrets.OLLAMA_API_KEY }}
     steps:
       - uses: actions/checkout@v4
         with:
@@ -61,10 +58,9 @@ jobs:
           echo "Models to publish: $models"
           echo "models=$models" >> "$GITHUB_OUTPUT"
 
-      - name: Check the secrets
+      - name: Check the secret
         run: |
           [ -n "$OLLAMA_KEY" ] || { echo "::error::Add the OLLAMA_KEY secret (contents of ~/.ollama/id_ed25519 from a machine signed in to the Tromset account)."; exit 1; }
-          [ -n "$OLLAMA_API_KEY" ] || echo "::warning::OLLAMA_API_KEY is not set: the model pages will not be updated."
 
       - name: Free disk space for the base models
         if: steps.pick.outputs.models != 'none'
@@ -96,16 +92,6 @@ jobs:
           # shellcheck disable=SC2086
           if [ "$MODELS" = all ]; then sh scripts/publish.sh Tromset; else sh scripts/publish.sh Tromset $MODELS; fi
 
-      - name: Update the model pages
-        if: env.OLLAMA_API_KEY != ''
-        run: |
-          if sh scripts/update-ollama-pages.sh Tromset; then
-            echo "Model pages updated from docs/ollama/." >> "$GITHUB_STEP_SUMMARY"
-          else
-            echo "::warning::Some ollama.com pages were not updated (see the log). Paste docs/ollama/<model>.md into the page by hand."
-            echo "Some ollama.com pages were not updated: paste docs/ollama/<model>.md by hand." >> "$GITHUB_STEP_SUMMARY"
-          fi
-
       - name: Remove the key
         if: always()
         run: rm -f ~/.ollama/id_ed25519 ~/.ollama/id_ed25519.pub

Documentation.md

@@ -352,23 +352,19 @@ With one attempt per problem, a gap of one problem is noise; the gap to the firs
 
 ## 8. Publishing on ollama.com
 
-Publishing is automatic. On every change to `main` that touches a Modelfile, a page in `docs/ollama/` or the publishing scripts, the [publish-ollama](.github/workflows/publish-ollama.yml) workflow:
+Publishing is automatic. On every change to `main` that touches a Modelfile or the publishing script, the [publish-ollama](.github/workflows/publish-ollama.yml) workflow rebuilds the changed models from their Modelfiles and pushes them as `Tromset/flycoder0.3`, `Tromset/flycoder0.3fast`, `Tromset/flycoder0.3pro`, `Tromset/flycoder0.3:lite` and `Tromset/flycoder0.3:router`.
 
-1. rebuilds the changed models from their Modelfiles and pushes them as `Tromset/flycoder0.3`, `Tromset/flycoder0.3fast`, `Tromset/flycoder0.3pro`, `Tromset/flycoder0.3:lite` and `Tromset/flycoder0.3:router`;
-2. updates each model page on ollama.com from `docs/ollama/<model>.md`.
-
-It can also be started by hand from the Actions tab (Run workflow), with a list of models or `none` to update only the pages.
+It can also be started by hand from the Actions tab (Run workflow), with a list of models or `all`.
 
 One-time setup, in the repository's Settings > Secrets and variables > Actions:
 
 | Secret | Value |
 |---|---|
 | `OLLAMA_KEY` | the private key `~/.ollama/id_ed25519` of a machine linked to the Tromset account (`ollama signin` on that machine first) |
-| `OLLAMA_API_KEY` | an API key created at https://ollama.com/settings/keys, used to update the pages |
 
-Updating a page uses the same request as the page's Edit button; Ollama does not document it. If it is refused, the workflow shows a warning and the page can still be pasted by hand.
+The model pages are not published automatically: ollama.com accepts page edits only from a signed-in browser and refuses API keys. After changing `docs/ollama/<model>.md`, open the model's page on ollama.com, click Edit and paste the file.
 
-By hand, from any machine signed in to the account: `sh scripts/publish.sh` (all models) or `sh scripts/publish.sh Tromset flycoder0.3fast` (one model), then `OLLAMA_API_KEY=... sh scripts/update-ollama-pages.sh`.
+By hand, from any machine signed in to the account: `sh scripts/publish.sh` (all models) or `sh scripts/publish.sh Tromset flycoder0.3fast` (one model).
 
 For a new version, update the version in `package.json`, `install.sh`, `scripts/publish.sh`, `scripts/publish-hf.sh`, the system prompt of the Modelfiles and the pages. `npm test` checks that they agree.
 
@@ -408,7 +404,7 @@ By hand: `pip install huggingface_hub`, then `HF_TOKEN=... sh scripts/publish-hf
 | Thinking takes too long | simple question handled in thinking mode | `--think=false` or `/set nothink` |
 | Endless repetition | temperature changed to a value that is too low | go back to the Modelfile's settings |
 | `npm run bench`: "Reference solution failed" | `python3` or `node` missing, or sandbox refused | `xcode-select --install`, Node 22, or `--no-sandbox` |
-| The publish workflow fails at "Check the secrets" | `OLLAMA_KEY` is missing | add it (section 8) |
+| The publish workflow fails at "Check the secret" | `OLLAMA_KEY` is missing | add it (section 8) |
 | publish-huggingface warns "HF_TOKEN is not set" and publishes nothing | `HF_TOKEN` is missing | add it (section 8.1) |
 
 ## 10. Development
@@ -422,8 +418,8 @@ By hand: `pip install huggingface_hub`, then `HF_TOKEN=... sh scripts/publish-hf
 | `brain/flybrain.mjs`, `brain/router.mjs` | FlyBrain server and routing rules |
 | `brain/eval-router.mjs`, `tests/fixtures/route-prompts.json` | router accuracy measurement |
 | `install.sh` | one-command installer |
-| `scripts/publish.sh`, `scripts/update-ollama-pages.sh` | publishing models and pages on ollama.com |
-| `docs/ollama/` | the ollama.com pages, one file per model |
+| `scripts/publish.sh` | publishing the models on ollama.com |
+| `docs/ollama/` | the ollama.com pages, one file per model, pasted by hand |
 | `docs/huggingface/`, `scripts/publish-hf.sh` | the Hugging Face model cards and their publisher |
 | `bench/bench.mjs`, `bench/problems.mjs` | benchmark and problems |
 | `bench/baselines/` | 0.1 profile, for comparison |

readme.md

@@ -63,7 +63,7 @@ Details: [Documentation.md, section 5.5](Documentation.md#55-flybrain-the-router
 
 ## Publishing
 
-Every change to `main` that touches a Modelfile or a page in `docs/ollama/` is published on ollama.com by the [publish-ollama](.github/workflows/publish-ollama.yml) workflow: models first, then their pages. It needs two repository secrets, `OLLAMA_KEY` and `OLLAMA_API_KEY` ([Documentation.md, section 8](Documentation.md#8-publishing-on-ollamacom)).
+Every change to `main` that touches a Modelfile is published on ollama.com by the [publish-ollama](.github/workflows/publish-ollama.yml) workflow. It needs the repository secret `OLLAMA_KEY` ([Documentation.md, section 8](Documentation.md#8-publishing-on-ollamacom)). The model pages are pasted by hand from `docs/ollama/`.
 
 Each publication is then mirrored on Hugging Face by the [publish-huggingface](.github/workflows/publish-huggingface.yml) workflow: the same GGUF weights, vision projector, system prompt and sampling settings, with a model card from `docs/huggingface/`, under [huggingface.co/Tromset](https://huggingface.co/Tromset). It needs the `HF_TOKEN` secret ([Documentation.md, section 8.1](Documentation.md#81-publishing-on-hugging-face)).
 
@@ -72,7 +72,7 @@ Each publication is then mirrored on Hugging Face by the [publish-huggingface](.
 - `Modelfile`, `Modelfile.fast`, `Modelfile.pro`: the three variants. `Modelfile.lite`, `Modelfile.router`: FlyBrain's experts.
 - `docs/ollama/`: the ollama.com page of each model. `docs/huggingface/`: the Hugging Face model card of each model.
 - `brain/`: the FlyBrain server and its routing rules.
-- `install.sh`: one-command install. `scripts/publish.sh`, `scripts/update-ollama-pages.sh`, `scripts/publish-hf.sh`: publishing.
+- `install.sh`: one-command install. `scripts/publish.sh`, `scripts/publish-hf.sh`: publishing.
 - `bench/`: quality and speed benchmark. `tests/`: `npm test`.
 - `training/`, `scripts/build_language_dataset.js`, `server/fly-language.js`: the former FlyBrain voice pipeline, independent of FlyCoder.
 

tests/modelfile.test.mjs

@@ -91,12 +91,11 @@ test('every published model has a short, up-to-date Ollama page', () => {
   }
 });
 
-test('the publish workflow runs the publisher and updates every page', () => {
+test('the publish workflow runs the publisher with the Tromset key', () => {
   const workflow = read('.github/workflows/publish-ollama.yml');
   assert.match(workflow, /branches: \[main\]/);
   assert.match(workflow, /sh scripts\/publish\.sh Tromset/);
-  for (const v of VARIANTS.filter(v => v.page)) assert.ok(workflow.includes(v.page.split('/').pop()) || workflow.includes('docs/ollama/'), `workflow uploads ${v.page}`);
-  for (const secret of ['OLLAMA_KEY', 'OLLAMA_API_KEY']) assert.ok(workflow.includes(`secrets.${secret}`), `workflow uses ${secret}`);
+  assert.ok(workflow.includes('secrets.OLLAMA_KEY'), 'workflow uses OLLAMA_KEY');
 });
 
 test('every published model has a Hugging Face model card and is mirrored by the Hugging Face publisher', () => {
@@ -117,5 +116,5 @@ test('every published model has a Hugging Face model card and is mirrored by the
 });
 
 test('shell scripts parse with POSIX sh', () => {
-  for (const file of ['install.sh', 'scripts/publish.sh', 'scripts/publish-hf.sh', 'scripts/update-ollama-pages.sh']) execFileSync('sh', ['-n', new URL(`../${file}`, import.meta.url).pathname]);
+  for (const file of ['install.sh', 'scripts/publish.sh', 'scripts/publish-hf.sh']) execFileSync('sh', ['-n', new URL(`../${file}`, import.meta.url).pathname]);
 });